Rigid, conjugated, fluoresceinated thymidine triphosphates: Syntheses and polymerase mediated incorporation into DNA analogues

Article abstract of DOI:10.1002/chem.200304944

Syntheses of a unique set of energy transfer dye labeled nucleoside triphosphates, compounds 1-3, are described. Attempts to prepare these compounds were only successful if the triphosphorylation reaction was performed before coupling the dye to the nucle

Full text of DOI:10.1002/chem.200304944

K. Burgess et al.
Chem. Eur. J. 2003, 9, 4603 4610
DOI: 10.1002/chem.200304944
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4603

FULL PAPER
Rigid, Conjugated, Fluoresceinated Thymidine Triphosphates: Syntheses
and Polymerase Mediated Incorporation into DNA Analogues
LarsH. Thoresen, [a] Guan-Sheng Jiao,[a] Wade C. Haaland,[b] Michael L. Metzker,[b]
and Kevin Burgess*[a]
Abstract: Syntheses of a unique set of
energy transfer dye labeled nucleoside
triphosphates, compounds 1 3, are de-
scribed. Attempts to prepare these com-
pounds were only successful if the tri-
phosphorylation reaction was per-
formed before coupling the dye to the
nucleobase, and not the other way
around. Compounds were prepared as
both the 2'-deoxy (a) and 2',3'-dideoxy-
(b) forms. They feature progressively
longer rigid conjugated linkers connect-
ing the nucleobase and the hydroxyxan-
thone moiety. UV spectra of the parent
nucleosides 12 14 show that as the
length of the linker increases so does
the absorption of the donor in the 320
330 nm region, but with relatively little
red-shift of the maxima. Fluorescence
spectra of the same compounds show
that radiation in the 320 330 nm region
results in predominant emission from
the fluorescein. When the linker is
irradiated at 320 nm, the only significant
emission observed corresponds to the
hydroxyxanthone part of the molecules
at 520 nm; this corresponds to an effec-
tive Stokes× shift of 200 nm. As the
absorption at 320 330 nm by the linker
increases with length, so does the inten-
sity of the fluorescein emission. A gel
assay was used to gauge relative incor-
poration efficiencies of compounds 1 3,
dTTP, ddTTP, and 6-TAMRA-ddTTP.
Throughout, the thermostable polymer-
ase TaqFS was used, as it is the one most
widely applied in high throughput DNA
sequencing. This assay showed that only
compounds 3 were incorporated effi-
ciently; these have the longest linkers.
Of these, the 2'-deoxy nucleoside 3a was
incorporated and did not prevent the
polymerase from extending the chain
further. The 2',3'-dideoxy nucleoside 3b
was incorporated only about 430 times
less efficiently than ddTTP under the
same conditions, and caused chain ter-
mination. The implications of these
studies on modified sequencing proto-
cols are discussed.
Keywords: DNA sequencing ¥ dyes/
pigments ¥ fluorescence spectrosco-
py ¥ fluorescent probes ¥ UV/Vis
spectroscopy
side triphosphates used in DNA sequencing are typical
examples of this, such as 6-TAMRA-ddTTP A; Figure 1.[1]
Rigid, non-aliphatic linkers have been investigated, but only
to link other labels. The types of probes that have been
connected in this way include weakly fluorescent, hydro-
Introduction
Many situations in biotechnology and biomedicine require
that DNA be labeled with highly fluorescent labels. Almost
invariably, this is done by connecting nucleosides to labels like
fluorescein and rhodamine via aliphatic linkers. The nucleo-
10]
phobic groups (e.g. B and C),[2 4] metal complexes (e.g. D),[5
EPR spin labels (E),[11] and electrochemically conspicuous
groups (F)[12] to nucleobases, but never the types of fluores-
cent labels that give the sensitivity required for application in
DNA sequencing and some areas of diagnostics.
[a] Prof. K. Burgess, Dr. L. H. Thoresen, Dr. G.-S. Jiao
Texas A & M University, Chemistry Department
P.O. Box 30012, College Station, TX 77842 (USA)
Fax : (‡1)979-845-8839
The lack of research into nucleobases labeled with strongly
fluorescent compounds via rigid conjugated linkers is un-
fortunate because systems like this could be useful. Relatively
inflexible linkers have the potential to limit intercalation or
other non-covalent interactions of the dye with the DNA
framework. They could also be used to keep the dye oriented
away from the enzyme active site of polymerases. Moreover,
conjugated linkers could electronically couple with the
fluorescent label and the nucleobase in ways that affect their
UV absorption and fluorescent properties in potentially
useful ways. For instance, this coupling could increase the
E-mail: Burgess@tamu.edu
[b] Dr. W. C. Haaland, Prof. M. L. Metzker
Department of Molecular and Human Genetics
and Human Genetics and Human Genome Sequencing Center
Baylor College of Medicine, Houston, TX 77030 (USA)
Supporting information for this article is available on the WWW under
http://www.chemeurj.org or from the author. Complete experimental
details for the preparation of the compounds, the spectroscopic
measurements (with some additional UV/fluorescence data for com-
pounds 12 14), and protocols for the polymerase mediated incorpo-
ration studies.
4604
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/chem.200304944
Chem. Eur. J. 2003, 9, 4603 4610

Fluoresceinated Nucleosides
4603 4610
cein fragments are connected to thymidine via progressively
longer, alkyne-based linkers. The linkers in 1 3 moves the
relatively bulky fluorescein unit further away from the
Figure 1. Highly fluorescent labels linked to nucleobases via flexible
linkers, as in 6-TAMRA-dTTP, are the norm in DNA sequencing protocols.
For other applications, rigid conjugated linkers have been employed to link
weakly fluorescent groups, redox, EPR, and electrochemical probes, but
mostly to nucleotides for incorporation into synthetic DNA and never for
DNA sequencing.
nucleoside base in defined increments. We anticipated that
increasing the linker conjugation would influence the spectral
properties of the dye, and the ease of incorporation of the
triphosphates. Consequently, this paper also describes some
fundamental fluorescence properties of the parent nucleo-
sides, and tests for their incorporation by the thermostable
DNA polymerase TaqFS. TaqFS was chosen for this study to
test compounds 1 3 because of its wide use in fluorescent
dye-terminator based DNA sequencing.[1, 15, 16] Both the 2',3'-
dideoxy (b) and the 2'-deoxy forms (a) were assayed, the
latter to gauge for the possibility of multiple incorporation.
UV absorption of the labeled nucleoside so that the fluores-
cence intensity of the label is increased. Alternatively,
conjugated linkers might be designed to absorb in UV regions
where it could be advantageous to excite the fluorescent
probe, but in which the probe without the linker has relatively
low extinction coefficients. Finally, attaching nucleobases to
fluorescent probes via conjugated linkers facilitates fast
fluorescence energy transfer between them, and potentially
provides a probe for energy transfer to that base from
neighboring nucleotides.[13, 14]
For the reasons described above, we have begun to study
nucleobases connected to highly fluorescent labels via rigid
conjugated linkers. This paper describes syntheses of deoxy-
and dideoxynucleoside triphosphates 1 3 in which fluores-
Synthesis of the modified thymidine triphosphates 1 3:
Scheme 1 describes how the fluorescein derivatives required
for this work were made from 5-iodofluorescein 4. Com-
pounds 4 and 5, in regioisomerically pure form, were prepared
Chem. Eur. J. 2003, 9, 4603 4610
www.chemeurj.org
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4605

K. Burgess et al.
FULL PAPER
via fractional crystallization of a mixture of 5- and 6-iodo-
fluorescein from acetic anhydride.[17] Acylation and Sonoga-
shira coupling[18] of this material with trimethylsilylethyne
gave the protected alkyne 6 and the terminal alkyne 7 after
deprotection.
The critical fluorescein-containing fragment 9 for prepara-
tion of compounds 2 was formed via a Suzuki coupling,[19] as
shown in Scheme 1b; fortunately, desilylation and hydrolysis
of the acetate protecting group also occurs in this step.
Synthesis of the diyne 11, a key fragment in the formation of
compounds 3, was made via a Sonogashira/deprotection
sequence as shown in Scheme 1c.
To obtain the target materials, compounds 7, 9, and 11 must
be coupled to 5-iodo-U, and the 5'-hydroxyl group must
converted to a triphosphate. Initially, we tried to achieve this
via a route in which the coupling preceded the triphosphor-
ylation. After much experimentation, and a considerable
amount of time, we concluded that it is extremely hard to
phosphorylate nucleosides with large, reasonably lypophilic
dyes attached. There are probably several reasons for this, but
paramount are solubility issues that impact purification using
conventional ion exchange chromatography on Sephadex or
various HPLC column purifications. The compounds may
aggregate under these conditions and stick to the support.
Lack of 3'/5'-selectivity in the triphosphorylation protocols[20]
may also have been a factor contributing to the difficulties
with this approach. This set of experiments did, however,
provide samples of the nucleoside intermediates 12 14 and
these were useful for spectroscopic studies (see below).
The successful approach to the target triphosphates 1
3
was to form 2'-deoxy-5-iodouridine triphosphate, and 2',3'-
dideoxy-5-iodouridine triphosphate (prepared via the proto-
col outlined in the Supporting Information), then coupling the
dye fragments to these. These triphosphates are only soluble
in aqueous media, so it was important to find a water-soluble
catalyst for this transformation. Sonogashira couplings in
aqueous media have been performed using preformed
[Pd(PPh2Ar)4] where Arˆ 3-(NaO3SC6H4),[21] and by mixing
Pd(OAc)2 and PAr3 in aqueous buffer and ™aging∫ the
solution (Method A in the Experimental Section).[22] That
Scheme 1. Preparation of the fluorescein-derived synthons 7, 9, and 11.
4606
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2003, 9, 4603 4610

Fluoresceinated Nucleosides
4603 4610
proved to be less convenient than preparing a crude sample of
catalyst by reacting Na2PdCl4, NaBH4, H2O with PAr3 at
258C, isolating the solid material, then using this in the
coupling reaction (Method B). Lithium salts of the phos-
phates were used as the substrates in the coupling process, the
products were isolated as triethylammonium salts after HPLC
purification (Scheme 2).
Scheme 2. Coupling of triphosphates with the dye fragments 7, 9, and 11 to
give the target products 1 3.
UV absorption and fluorescence spectra of nucleosides 12
14: Figure 2 shows the absorption and fluorescence spectra for
nucleosides 12 14 corresponding to the triphosphates 1 3.
The dye-nucleosides have absorption maxima at 320 330 and
492 nm, probably corresponding to uptake of radiation by the
nucleobase in conjugation with its 5-substituent, and the
fluorescein, as illustrated in Figure 2 for compound 12
(though similar diagrams could be drawn for 13 and 14).
Experiments with control compounds show that the ab-
sorption spectra of these molecules resemble that which
would be obtained simply by adding their donor and acceptor
parts (a typical overlay of control compounds is given in the
Supporting Information). This is indicative of a lack of
conjugation between them, reflecting a twist between the
donor and acceptor planes imposed by steric effects, as
implied in Figure 2a. As the donor part of the molecules
becomes more extended on going from 12 14, so the
absorption in the 320 330 nm region becomes more intense.
This is consistent with observations for poly- or oligo-phenyl-
ethyne materials {(-C6H4CC-)n}, where the UV absorption
becomes stronger in this region, but with little shift to the red,
as more repeating units are added.[23 26] No emission from the
donor part of these molecules was detected in their fluo-
rescence spectra. Instead, rapid energy transfer seems to
Figure 2. a) Representative assignment of donor and acceptor parts in one
of the nucleosides 12; b) absorption spectra (all at 10 mm); c) fluorescence
spectra of the nucleosides in 3.2 mm phosphate buffer at pH 7.2 at
equimolar concentrations (all at 1.0 mm).
prevail and fluorescence is observed from the hydroxyxan-
thanone acceptor unit. The fluorescence intensity of the
cassettes when irradiated at 320 330 nm nearly doubles from
compound 12 to 14 due to the increased UVabsorption by the
latter.
Incorporation of the modified thymidine triphosphates 1 3:
The performance of compounds 1 3 was tested using the
™oligo-template assay∫ using TaqFS as previously descri-
bed,[27] and was analyzed using an ABI model 377 DNA
sequencer. Here, the BODIPY-R6G labeled R931 universal
sequencing primer[28] was used to determine incorporation of
the fluorescein derivatives 1 3, 6-TAMRA-ddTTP, or unla-
beled ddTTP or dTTP. The sequencer run module was
Chem. Eur. J. 2003, 9, 4603 4610
www.chemeurj.org
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4607

K. Burgess et al.
FULL PAPER
modified to collect fluorescent blue, green, and yellow
fluorescent light to discriminate signals emitted by the
fluorescein derivatives 1 3, BODIPY-R6G, and 6-TAMRA
dyes, respectively. Moreover, we have previously shown that
incorporation of modified-dNTPs tend to give DNA frag-
ments with slower electrophoretic mobilities than natural
substrates because of their high molecular weights.[27] The
attachment of fluorescein moieties to dTTP and ddTTP
caused similar mobility effects. Thus, the assignment of
incorporation events was based on the color and mobility
pattern of termination products relative to the unlabeled
ddTTP and dTTP controls.
ent incorporation in the oligo-template assay (Figure 3).
Fluorescent intensities of incorporated and unincorporated
termination bands were quantitated and expressed as the
percentage incorporated (I%). Tests for incorporation at
different concentrations of the ddNTPs (performed in tripli-
cate) are plotted in Figure 4. The relative order of preferred
incorporation for TaqFS was compounds ddTTP > 6-TAM-
RA-ddTTP > 3b > 2b > 1b.
Each of the substrates 1 3 was tested at a series of
concentration ranges to determine an IC50 value (effective
concentration at 50% incorporation) in the absence of the
corresponding natural nucleotide. Both the deoxy- and
dideoxy-thymidine analogues (a and b) were tested individ-
ually and compared with dTTP and ddTTP to assess TaqFS
bias incorporation effects. For the dideoxy-compounds (b
series), there are two possible outcomes: the compounds
could be incorporated causing chain termination or they are
not recognized by TaqFS. For the latter case, DNA synthesis is
halted in the absence of natural thymidine, shown as ™unin-
corporated∫ in Figure 3. Incorporation of the deoxy-com-
pounds (a series) would cause continued DNA synthesis or
™read-through∫ to the ™incorporated∫ end-point. We have
also observed chain termination of series a compounds and
hypothesize their incorporation distorts the DNA substrate
for subsequent DNA synthesis. Alternatively, TaqFS might
not recognize the substrate. Incorporation data for com-
pounds 2b and 3b are shown in Figure 3.
Figure 4. Incorporation curves for ddTTP, 6-TAMRA-ddTTP, and com-
pounds 3b and 2b. Dilution series for each compound were performed in
triplicate: the range of standard deviations were 0.2 4.7% for ddTTP, 1.6
15.6% for 6-TAMRA-ddTTP, 0.5 10.7% for 3b, and 0.8 6.8% for 2b.
From the titration plots, IC50 values were derived for
compounds 2b, 3a, and 3b, ddTTP, dTTP, and 6-TAMRA-
ddTTP (Table 1). To compare the relative performance of
nucleotide pairs, incorporation ratios (IR) were determined.
These experiments show that TaqFS prefers ddTTP to dTTP
or 6-TAMRA-ddTTP approximately 7.6-fold and 18-fold,
respectively. The IC50 values for compounds 2b and 3b,
however, suggested a greater bias against incorporation
compared to ddTTP (ꢀ4300-fold and ꢀ430-fold, respective-
ly). The 10-fold improvement of compound 3b over com-
pound 2b can most probably be attributed to the additional
length of the linker. Compared with 6-TAMRA-ddTTP,
compound 3b showed less bias against incorporation (ꢀ25-
fold) than other reported fluorescein-labeled dTTP analogues
(ꢀ50-fold). TaqFS seems to have an inherent bias in favor of
rhodamines as opposed to fluoresceins,[1] but this is less
evident for 3b.
The IC50 values determined for the deoxythymidine com-
pound 3a and its corresponding dideoxy-analogue 3b were
both 0.16 mm; these values are less than for the compounds 1
and 2 suggesting that the longer nucleobase linker dye
structure exerts a major influence on TaqFS incorporation.
DNA polymerase TaqFS showed less preference for dTTP
over the deoxythymidine analogue 3a than it did to ddTTP
over the dideoxythymidine analogue 3b (IR ˆꢀ 57-fold). The
oligo-template assay also showed that compound 3a was
Figure 3. Incorporation of compounds 3b and 2b by TaqFS DNA
polymerase. All reactions contained TaqFS, BODIPY-R6G labeled R931
primer, oligo-template and reaction buffer. Lane 1 contained only the
primer; lanes 2 18 contained 50 nm each of dCTP, dATP, and ddGTP.
Additionally, lanes 3 and 4 contained 10 nm ddTTP or 10 nm dTTP,
respectively and lanes 5 11 contained 3b, lanes 12 18 contained 2b at
500, 250, 100, 50, 25, 10, or 5 nm, respectively.
The linker had a profound effect on the incorporation of
compounds 1 3. For compounds 1a and b, there was no
evidence of incorporation (data not shown). On the other
hand, compounds 2b and 3b showed concentration depend-
4608
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2003, 9, 4603 4610

Fluoresceinated Nucleosides
4603 4610
Table 1. Summary of IC50 values and incorporation ratios (IR) for test
compounds 1 3 and controls dTTP, ddTTP, and 6-TAMRA-ddTTP.
acceptor dye in compounds 1 and 2 is too close to the
nucleobase, and TaqFS does not tolerate them well. Com-
pound 3, however, seems to have dimensions that are close to
the minimum separation which this polymerase needs for
efficient incorporation.
Compounds
IC50 [nm][a]
IR values
dTTP
2.8
ddTTP
1a
0.37
NT
[dTTP]/[ddTTP] ˆ 7.6
The prognosis for future work in this area seems good.
Modified nucleobase triphosphates with rigid conjugated
linkers that are even longer than those in 3 will absorb even
more strongly in the 320 330 nm range. Such compounds
could be designed to relay this energy to the terminal dye with
very large effective Stokes× shifts and result in increased
fluorescence intensities. The triphosphates should tend to be
incorporated more efficiently by polymerase enzymes like
TaqFS, and the rigid separation of the dye from the polymer-
ase active site should also increase the range of fluorescent
label structures that these enzymes will tolerate. One of the
reasons for the dominance of rhodamine-based dye-termina-
tor chemistries in DNA sequencing, for instance, is that other
dye types tend not be incorporated as efficiently.[1] Rigid
active site-dye separation could alleviate this restriction. In
DNA sequencing methodologies, this is of immediate interest
because there are several advantages to fluorescein-based
terminator chemistries that could be exploited if they were
incorporated more efficiently. Specifically, dye mobility
patterns are more predictable when fluorescein dyes are
used, hence they can be corrected more accurately, and the
unincorporated and extension by-products can be removed
more readily. Moreover, if single molecule detection methods
do become practical for DNA sequencing, then the chemical
nature of the linker moiety may be more critical. Single
molecule methods may be based on incorporation of labeled
terminators, or on multiple incorporations of labeled 2'-
deoxynucleoside triphosphates.[29] In the latter case, optimi-
zation of the linker geometry for clean and efficient incorpo-
ration may be essential. Similar considerations apply to some
diagnostic protocols featuring incorporation of labeled bases
in the PCR reaction.[30] Overall, the prospects for research in
this area appear to be as bright as the labels developed.
1b
2a
2b
3a
NT
NS
1.6 Â 103
[2b]/[ddTTP] ˆ 4.3  103
[3a]/[dTTP] ˆ 57
160
3b
160
6.5
[3b]/[ddTTP] ˆ 4.3  102
[6-TAMRA]/[ddTTP] ˆ 18
6-TAMRA-ddTTP
[a] ™NT∫ means no termination event was observed up to
concentration of 5 mm, and ™NS∫ means the observed termination products
were not sufficient to determine an IC50 value.
a final
incorporated and extended the R931 primer to the end-point
nucleotide, although a second minor product was revealed,
which corresponded to a thymidine termination product (data
not shown); that is, 3a was incorporated and only partially
impeded subsequent DNA synthesis, hence the enzyme was
able to ™read through∫ to the end of the strand.
Conclusion
The synthetic chemistry described here shows how fluorescein
derivatives can be functionalized with phenylethyne linkers to
give nucleobase fluorescein energy transfer cassettes. These
particular linkers absorption in the 320 330 nm range, and
excitation in that region leads to efficient fluorescence
emission at the fluorescein dye. To understand the implica-
tions of these observations in areas where the intensity of
fluorescence detection is critical, it is important to keep in
mind the differences between the concepts of quantum yield
and fluorescence intensity. Fluorescence intensity is the
critical factor that determines how much radiation is emitted
from a fluorescent probe. High quantum yields are necessary
to obtain strong fluorescence intensities, but they are not the
sole criteria: extinction coefficients of the dye at the wave-
length of the excitation source are also critical. For a dye like
fluorescein, it is not possible to significantly increase the
quantum yield because it is near unity. However, by con-
jugating other groups onto fluorescein it is possible to increase
the amount of energy that is channeled into a dye and tune
the wavelengths at which the exciting photons are best
absorbed.
This study illustrates that rigid conjugated linkers have
absorptions that increase with linker length, that is the longer
the linker the more energy that is harvested and fed into the
dye. Predictably, long conjugated linkers have higher extinc-
tion coefficients than shorter ones. They also have absorption
lmax values that shift to the red, away from the intense
absorption of DNA bases that would otherwise compete for
the incident photons (ca. 250 300 nm). Our work also shows
that, fortunately, long conjugated linkers are also desirable
from the perspective of enzyme incorporation. While the
Acknowledgement
Use of the TAMU/LBMS-Applications Laboratory directed by Dr. Shane
Tichy is acknowledged. Support for this work was provided by The
National Institutes of Health (HG 01745) and by The Robert A. Welch
Foundation. Instrumentation was supplied by the NIH Laser Resource
Grant P41RR03148.
[1] J. W. Brandis, Nucleic Acids Res. 1999, 27, 1912.
[2] V. A. Korshun, I. A. Prokhorenko, S. V. Gontarev, M. V. Skoroboga-
tyi, K. V. Balakin, E. V. Manasova, A. D. Malakhov, Y. A. Berlin,
Nucleosides Nucleotides 1997, 16, 1461.
[3] V. A. Korshun, E. V. Manasova, K. V. Balakin, A. D. Malakhov, A. V.
Perepelov, T. A. Sokolova, Y. A. Berlin, Nucleosides Nucleotides 1998,
17, 1809.
[4] E. V. Malakhova, A. D. Malakhov, S. V. Kuznitsova, O. P. Varnavskii,
A. P. Kadutskii, D. T. Kozhich, V. A. Korshun, Y. A. Berlin, Bioorg.
Khim. 1998, 24, 688.
[5] D. J. Hurley, Y. Tor, J. Am. Chem. Soc. 2002, 124, 3749.
[6] H. Weizman, Y. Tor, J. Am. Chem. Soc. 2002, 124, 1568.
[7] M. T. Tierney, M. W. Grinstaff, Org. Lett. 2000, 2, 3413.
Chem. Eur. J. 2003, 9, 4603 4610
www.chemeurj.org
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4609

K. Burgess et al.
FULL PAPER
[8] S. I. Khan, A. E. Beilstein, M. W. Grinstaff, Inorg. Chem. 1999, 38,
418.
[9] S. I. Khan, A. E. Beilstein, M. T. Tierney, M. Sykora, M. W. Grinstaff,
Inorg. Chem. 1999, 38, 5999.
[10] S. I. Khan, M. W. Grinstaff, J. Am. Chem. Soc. 1999, 121, 4704.
[11] A. Spaltenstein, B. H. Robinson, P. B. Hopkins, J. Am. Chem. Soc.
1988, 110, 1299.
[12] C. J. Yu, H. Yowanto, Y. Wan, T. J. Meade, Y. Chong, M. Strong, L. H.
Donilon, J. F. Kayyem, M. Gozin, G. F. Blackburn, J. Am. Chem. Soc.
2000, 122, 6767.
[19] N. Miyaura, A. Suzuki, Chem. Rev. 1995, 95, 2457.
[20] K. Burgess, D. Cook, Chem. Rev. 2000, 100, 2047.
[21] A. Casalnuovo, J. C. Calabrese, J. Am. Chem. Soc. 1990, 112, 4324.
[22] D. T. Bong, M. R. Ghadiri, Org. Lett. 2001, 3, 2509.
[23] J. S. Schumm, D. L. Pearson, J. M. Tour, Angew. Chem. 1994, 106,
1445; Angew. Chem. Int. Engl. Ed. 1994, 33, 1360.
[24] T. M. Swager, C. J. Gil, M. S. Wrighton, J. Phys. Chem. 1995, 99, 4886.
[25] C. Devadoss, P. Bharathi, J. S. Moore, J. Am. Chem. Soc. 1996, 118,
9635.
[26] C. Weder, M. S. Wrighton, Macromolecules 1996, 29, 5157.
[27] M. L. Metzker, R. Raghavachari, S. Richards, S. E. Jacutin, A.
Civitello, K. Burgess, R. A. Gibbs, Nucleic Acids Res. 1994, 22, 4259.
[28] M. L. Metzker, J. Lu, R. A. Gibbs, Science 1996, 271, 1420.
[29] L. M. Davis, F. R. Fairfield, C. A. Harger, J. H. Jett, R. A. Keller, J. H.
Hahn, L. A. Krakowski, B. L. Marrone, J. C. Martin, Genet. Anal.
Tech. Appl. 1991, 8, 1.
[13] S. Georghiou, S. Zhu, R. Weidner, C. R. Huang, G. Ge, J. Biomol.
Struct. Dyn. 1990, 8, 657.
[14] D.-G. Xu, T. M. Nordlund, Biophys. J. 2000, 78, 1042.
[15] B. B. Rosenblum, L. G. Lee, S. L. Spurgeon, S. H. Khan, S. M.
Menchen, C. R. Heiner, S. M. Chen, Nucleic Acids Res. 1997, 25, 4500.
[16] L. G. Lee, S. L. Spurgeon, C. R. Heiner, S. C. Benson, B. B. Rose-
nblum, S. M. Menchen, R. J. Graham, A. Constantinescu, K. G.
Upadhya, J. M. Cassel, Nucleic Acids Res. 1997, 25, 2816.
[17] G.-S. Jiao, J. W. Han, K. Burgess, unpublished results.
[18] K. Sonogashira, Y. Tohda, N. Hagihara, Tetrahedron Lett. 1975, 16,
4467.
[30] Z. Zhu, J. Chao, H. Yu, A. S. Waggoner, Nucleic Acids Res. 1994, 22,
3418.
Received: March 13, 2003 [F4944]
4610
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2003, 9, 4603 4610